There are a number of biological and medical applications that are currently impractical due to limitations in cell and particle analysis technology. Examples of such biological applications include battle field monitoring of known airborne toxins, as well as the monitoring of cultured cells to detect the presence of both known and unknown toxins. Medical applications include non-invasive prenatal genetic testing and routine cancer screening via the detection and analysis of rare cells (i.e., low rate of occurrence) in peripheral blood. All of these applications require an analysis system with the following principal characteristics:
1. high speed measurement; PA1 2. the ability to process very large or continuous samples; PA1 3. high spectral resolution and bandwidth; PA1 4. good spatial resolution; PA1 5. high sensitivity; and PA1 6. low measurement variation.
In prenatal testing, the target cells are fetal cells that cross the placental barrier into the mother's bloodstream. In cancer screening, the target cells are sloughed into the bloodstream from nascent cancerous tumors. In both of these applications of this technology, the target cells may be present in the blood at concentrations of one to five cells per billion. This concentration yields approximately 20-100 target cells in a typical 20 ml blood sample. The extreme rarity of the targeted cells demands that any detection and analysis system employed in these applications be capable of processing an enriched sample of approximately 100 million cells within a few hours, corresponding to a minimum throughput of 10,000 cells per second. Cell processing includes the determination of cellular morphology parameters such as overall size, nuclear size, nuclear shape and optical density, the detection and characterization of numerous fluorescent markers and FISH probes, the quantification of the total amount of DNA in the nucleus, and the detection of other cellular components such as fetal hemoglobin. To accomplish these processing tasks, the system must be able to collect cell images with a spatial resolution of approximately 1 micron. Likewise, the system must have high spectral resolution and bandwidth to differentiate four or more fluorescent colors. Since some probes may label important cellular features with only a few thousand fluorescent molecules, the system must have high sensitivity and good measurement consistency to differentiate very weak signals.
The predominant research laboratory protocols for non-invasive prenatal diagnosis employ a complex series of process steps that include gradient centrifugation to remove unnucleated cells, high-speed cell sorting for fetal cell enrichment, and fluorescence microscopy for fetal cell identification and genetic analysis. These protocols often yield little or no fetal cells for analysis, because a fraction of the fetal culls are lost at each step of the protocol. Nevertheless, the protocols cannot be simplified because of limitations in existing analysis technology. Ideally, fetal cell identification and analysis would be performed in a few hours by a high speed cell sorter having the necessary speed and sample handling capacity. This ideal is not possible with conventional systems, because conventional cell sorters lack the necessary imaging abilities, sensitivity, and repeatability to reliably identify fetal cells and enumerate the number and color of FISH probes used to make the diagnosis. Therefore, under current protocols, cells must be sorted onto slides and examined using fluorescence microscopy to establish their fetal origin and make a genetic diagnosis. The combination of low fetal cell yields and lengthy processing times precludes the clinical application of non-invasive fetal testing with existing technology.
No technology prior to the present invention incorporates all six of the principal characteristics of a viable fetal cell or cancer analysis system. In the prior art, there have been advances that might be applied to these applications, but significant limitations still remain.
A paper published by Ong et al. [Anal. Quant. Cytol. Histol., 9(5):375-82] describes the use of a time-delay and integration (TDI) detector in an imaging flow cytometer. A TDI detector is any pixellated device in which the signal produced in response to radiation directed at the device can be caused to move in a controlled fashion. Typically, the pixels of a TDI detector are arranged in rows and columns, and the signal is moved from row to row in synchrony with a moving image projected onto the device, allowing an extended integration time without blurring. The approach disclosed by Ong et al. advanced the art by addressing the need for spatial resolution and high sensitivity for cells in flow. However, this approach does not address the remaining principal characteristics. The authors of this paper cite an operating speed of 10 cells per second and a theoretical speed limitation of 500 cells per second, which is at least an order of magnitude slower than is required for non-invasive fetal testing. In addition, the system has no spectral resolution; laser scatter and fluorescence light are collected by the imaging system indiscriminately.
In more recent developments, U.S. Pat. No. 5,644,388 discloses an alternative approach to an imaging flow cytometer. The patent discloses the use of a frame-based image collection approach in which a video camera views cells in flow, in a freeze frame fashion. This method requires the image collection system to be synchronized with the presence of cells in the imaging area, unlike the case of TDI, wherein the detector readout rate is synchronized with the velocity of the cells. When a cell is imaged with the frame-based method, the integration period must be very short to prevent blurring. A short integration time is achieved either with a strobed light source, or a continuous light source combined with a shuttered detector. In either case, the short integration time reduces the signal-to-noise ratio and the ultimate sensitivity of the approach. Further, frame-based cameras require time to transfer data out of the camera, during which no images arc acquired, and cells of interest can escape detection. Finally, like the work of Ong et al, this patent makes no provisions for acquiring data over a large spectral bandwidth and with sufficient spectral resolution to simultaneously resolve numerous and differently colored fluorescent probes and FISH spots.
Spectral discrimination is addressed in U.S. Pat. No. 5,422,712, in which the spectra of particles suspended in a fluid are collected as the particles flow through a detection region. However, there is no spatial representation of the object in the system disclosed in this patent, because the object is defocussed at the detector. In this system, light is collected from the object and an image is created at an intermediate aperture. The light continues through the aperture to a spectral dispersing element, which disperses the light spectrally along the axis of flow. The dispersed light is applied to an image intensifier in which it is amplified, and the light signal output from the image intensifier is finally directed to frame-based detector. At the intermediate aperture, prior to spectral dispersion, the image represents the spatial distribution of light in object space. The spatial distribution is blurred as the light propagates past the image plane, through the spectral dispersing element and onto the image intensifier. Because there is no provision for re-imaging the intermediate aperture at the intensifier, the resulting signal distribution at the intensifier represents only the spectral distribution of the light and does not preserve the spatial distribution of the light from the object. The loss of spatial information limits the utility of the invention for applications such as fetal cell analysis. If multiple identical FISH spots are present in a cell, their spectra can be ascertained using this approach, but the number of spots cannot be determined. In addition, this approach disperses the wavelength spectrum parallel to the axis of flow. If two particles are illuminated in the flow axis, their spectra can overlap on the detector. To prevent this problem, the patent discloses that a very short illumination height in the flow axis is used. The short illumination height decreases integration time, which necessitates the use of the image intensifier. Further, the short illumination height limits throughput by preventing the simultaneous imaging of multiple cells in the flow axis.
Accordingly, it will be apparent that an improved technique is desired that resolves the limitations of the conventional approaches discussed above. It is expected that the new approach developed to address these problems in the prior art will also have application to the analysis of other types of moving objects besides cells and may be implemented in different configurations to meet the specific requirements of disparate applications of the technology.